1. Field of the Invention
The present invention relates to electrochemical cell probes and, more particularly, to probes that measure the solute contents of a variety of elements dissolved in liquid metals.
2. Brief Description of the Prior Art
Electrochemical cell (emf) probes are commonly used for in situ measurements of oxygen (O) levels in liquids, such as molten metals. An example includes the CELOX brand oxygen emf probe, available commercially from the Heraeus/Electro-Nite Co. Emf recordings, such as those recorded with the CELOX brand oxygen emf probe, are also used to determine the dissolved aluminum content in deoxidized steel, with the percent of dissolved aluminum in the steel bath being derived from the CELOX brand emf readings for oxygen activity and aluminum thermodynamic data compiled by Hultgren et al. (see "Selected Values of Thermodynamic Properties of Metals and Alloys", John Wiley, New York, 1974), on the assumption that the deoxidation product is pure Al.sub.2 O.sub.3. However, no detailed study has been made to evaluate the uncertainty limits of the calculated aluminum content.
Many of the commercial oxygen emf probes have three main deficiencies. First, the oxygen emf probes have a thin wall ZrO.sub.2 (MgO) electrolyte (.about.1 mm) and at very low oxygen contents, such as with aluminum deoxidation of steel, there is some electronic conduction in the ZrO.sub.2 (MgO) electrolyte. Consequently, the oxygen emf probe registers an emf reading that is somewhat higher than with an electrolyte where the electronic conduction is negligibly small. A second deficiency is that at high oxygen contents, there is rapid oxygen diffusion from the melt through the thin wall electrolyte, causing polarization of the reference electrode (a compound of Cr+Cr.sub.2 O.sub.3). This manner of polarization induces lower emf readings, and will therefore predict erroneous lower oxygen contents. It is for this reason that an empirical formulation of the emf versus ppm O relation is used for the CELOX brand emf reading. Finally, a third deficiency is that the oxygen emf probes do not directly detect the presence of solute contents other than oxygen in liquid metals.
In an effort to design probes that can detect dissolved elements other than oxygen, research has been conducted to develop a silicon emf probe for the in situ measuring of silicon (Si) levels in liquid metals (see M. Iwase, H. Abe, and H. Iritani, Steel Research 59, No. 10, pp. 433-437 (1988)). In general, stabilized zirconia electrolytes are coated with a metal silicate to define the oxygen chemical potential at the liquid iron-electrolyte interface, e.g., Mg.sub.2 SiO.sub.4, ZrSiO.sub.4, and Na.sub.2 Si.sub.2 ZrO.sub.2. However, because of the lack of thermodynamic data on all the components of the silicon emf probe, a rigorous theoretical interpretation of the measured silicon probe emf has not been possible. Instead, the percentage of silicon in hot metal versus the probe emf relation is derived by an empirical formulation based on plant tests. As an example, the experimental results obtained by Gomyo et al. (see Iron and Steelmaker, 1991, 18(7), 71; 1993, 20(3), 87) using Fe--C--Si melts at 1450.degree. C. and an (Mo+MoO.sub.2) reference electrode are in general accord with the results of measurements made in the blast furnace runner at the United States Steel Fairless Works, reported by McDowell and Clauss (see Ironmaking Conf. Proc., 49, 691, (1990)). The plant tests were made using a SILTEMP brand silicon emf probe available commercially from Leeds & Northrup, with a molybdenum (Mo+MoO.sub.2) reference electrode and a ZrO.sub.2 (MgO) electrolyte coated with an undisclosed silicate. In these tests, the hot metal temperatures were in the range of 1425.degree. C..+-.30.degree. C. At 0.1% of silicon in hot metal, the reproducibility is within .+-.0.04% Si. This increases to .+-.0.10% Si at 1.0% Si and to about .+-.0.15% Si at 1.2% Si.
In most cases, molten silicates and aluminosilicates have been used satisfactorily in silicon emf probes since the ionic transport number in these polymeric melts is close to unity and the solute content of liquid metal can be derived from the measured probe emf using the thermodynamic relation between the probe emf and the solute activity in the liquid metal. However, endeavors to develop emf probes that directly detect carbon (C), sulfur (S), nitrogen (N), phosphorous (P), aluminum (Al), silicon (Si), and chromium (Cr) in liquid metals have not succeeded. This is because the carbides, nitrides, and sulfides used as emf probe electrolytes do not have the desired properties for the ionic conduction of C, N, and S. These elements do dissolve in polymeric melts in ionic form (C.sup.2-, N.sup.3-, S.sup.2-) however.